MITHRIL MEASURED RECEIPTS
Measured receipts
These receipts are not an Artificial Analysis rank and not a BenchLeader place.
| Measurement | Score | Detail | Measured |
|---|---|---|---|
| Kept local todo completion | 1 | cold median 0.0002264160000000004 s. reflections 1. examined 3. USD 0. 0 model calls. 511 boolean vectors. | 2026-10-12T15:07:24.045Z |
| Latest unkept calibration | 1 | Score stays 1. This median is not the kept gate. cold median 0.00024149999999999495 s. | 2026-10-12T19:52:20.229Z |
| ITBench-AA public Scenario-11, alerts only | 0 | TP 0, FP 3, FN 2 | 2026-10-12T18:26:15Z |
| ITBench-AA public Scenario-11, topology | 0 | TP 0, FP 15, FN 2 | 2026-10-12T18:46:23Z |
| τ³-Banking public task_006, one trial | 0 | pass^1 0 | 2026-10-13T04:06:49.455176+09:00 |
The kept result still visits all 511 boolean vectors of length 0-8. Provider cost is USD 0 and model calls are 0. The local harness rate is not a BenchLeader place. Official Artificial Analysis ranks use their full task sets and procedures. These public slices are the runs that completed.